Eight creative ideas for more democracy

Voluntary citizens’ panels debate topical issues

Models are currently trending in which citizens can consult with one another and put forward proposals to either politicians or the general public. Various forms of deliberative and participatory democracy are being tried out in democratic countries. For example, the Genevan political scientist Nenad Stojanović launched the ‘Demoscan’ pilot project in Sion in 2019. It’s the first-ever citizens’ panel in Switzerland.

Twenty randomly selected individuals came together over two weekends to compile a citizens’ letter with the most important arguments for and against a popular initiative. This letter was then included with the official materials sent to all citizens when the initiative was put to a referendum. This project, meanwhile renamed ‘Pidem’, has since trialled similar voluntary mini-panels in places such as Prilly, Thalwil, Bellinzona and Geneva. The media response has been mixed. Those participating find it a great experience, though it remains uncertain just what impact their recommendations have on voters.

AI systems engaging in democratic debate

In democratic decision-making processes based on so-called deliberative principles, people work together to find solutions that are acceptable to everyone. “It’s in groups that we are generally well placed to do this in a way that leads to a specific outcome”, says Francesco Veri, a political scientist at the University of Zurich. In his view, to have a reasonable discussion, your arguments have to be linked coherently with the political measures you favour. “I need to be able to grasp why other people think differently – but without losing sight of my own position”. This is why he wanted to investigate whether the machine-generated reasoning of chatbots is comparable to how human beings consider the broader picture when they have a group discussion.

To this end, Veri has worked with Simon Niemeyer from Canberra to develop the Deliberative Reasoning Index (DRI). It enables them to measure whether collective deliberation can lead to a shared understanding of a topic. They got human participants to complete a questionnaire before and after town hall meetings. Then they presented the same topic and the same questions to various language models. Their results showed that while opinions, values and preferred measures are clearly linked together in human discourse, the decision-making processes of AI systems don’t necessarily exhibit the same coherence.

Veri then did a second experiment in which he got the chatbots to assume the role of a militant environmentalist, or a farmer affected by climate change. He found that the machines were indeed capable of generating appropriate arguments but failed to maintain a consistent stance across multiple rounds of questioning, nor were they able to prioritise corresponding political measures. “They are simply unable to demonstrate any stable, consistent structure for their arguments”, he says. As a result, he’s warning people against using AI as a substitute for population groups not represented in parliament, for example. “We shouldn’t be treating AI systems like political decision-makers”.

Everyone can make suggestions for their city district

In 2019, Lausanne became the first Swiss city to introduce a participatory budget. It was initially set up as a three-year pilot project and allocated between CHF 100,000 and 175,000 each year for regeneration projects across the city’s neighbourhoods. Under this so-called participatory budgeting scheme, all residents of the city are allowed to propose how to spend the money, regardless of their nationality or age. The city administration then assesses whether the proposals fall under the correct jurisdiction, whether they are feasible or not, and thereafter places them before the people in a referendum. The projects are ranked according to the number of votes they get and are funded in descending order until the sum given has been spent. Up to CHF 20,000 can be allocated for each individual proposal.

“One core priority for the city was that it should reach those people better who don’t normally participate in the political process”, says Daniel Kübler from the Centre for Democracy Studies Aarau (ZDA), which was tasked with evaluating the pilot phase. Lausanne was aware that this pilot would require extra, supporting measures. Interested parties who had no prior experience were offered workshops to help them formulate their ideas in ways more likely to result in success, and to develop their proposals properly. “If we hadn’t done that”, says Kübler, “we’d simply risk addressing yet again those same people who are already politically active and familiar with democratic processes”.

Kübler was responsible for evaluating the pilot project, alongside a fellow researcher, Marine Benli-Trichet. Their core finding was this: “You have to invite people to participate in politics. You have to empower them, rather than just wait for them to come to the process of their own accord”. For a long time, people in Switzerland have been sceptical of participatory budgets. People tend to assume that their country already has sufficient instruments for direct democracy. “But it’s not enough just to set up new channels for participation”, says Kübler. What’s more, he insists, it takes time to mobilise new sections of the population and to get them involved. Lausanne seems to be on the right path in this regard, for their participatory budget is now entering its eighth year. And every year, between 20 and 30 proposals are submitted.

Citizens’ Council offers recommendations to politicians

Back in 2019, Belgium became the first country in the world to establish a permanent ‘citizens’ dialogue’, which became known as the ‘East Belgian Model’. It consists of a Citizens’ Assembly and a Citizens’ Council. The Assembly is made up of 25 to 50 people who represent a diverse cross-section of society, but who are otherwise selected randomly. They come together every year over several Saturdays to discuss a specific topic and draw up recommendations for parliament, which is obliged to assess them. After the Assembly has completed its task, its members are asked if they’d like to become members of the Citizens’ Council, which has 24 members who commit themselves to serving for a full three-year term.

The Citizens’ Council convenes the Citizens’ Assembly and decides what issues it should debate. These topics may be proposed not only by the Council, but also by the Parliamentary Office and indeed by anyone belonging to the population of East Belgium. Once a year, the general public is invited to submit its own suggestions. The Citizens’ Council also monitors how the Assembly’s recommendations are implemented. Parliament is allowed to reject them, but it has to give good reasons. According to Cecilia Baglietto from the University of Louvain, that happens only rarely. Her university has been monitoring the process since it first began. When problems occur, she says, it’s usually because recommendations have been formulated in very open terms: “This leaves parliament a good deal of leeway when it comes to interpreting and implementing them”. As a result, the Citizens’ Council has begun engaging in a more frequent dialogue with smaller groups of politicians in order to discuss what the core issues are.

The biggest challenge of all, however, is long-term commitment. The Citizens’ Council in particular demands a big investment of time and energy. If people have little free time – thus Baglietto – or if their life circumstances change frequently, then it can be difficult for them to attend all the meetings. But it’s precisely the continuity of the Model that is its great strength, she says. Its long-term nature means it really helps to establish a proper relationship between the members of the Citizens’ Council and parliament.

How young people inform themselves about politics today

“We still know very little about how young people engage with political content online”, says Monika Waldis Weber, who is currently running a project on this very topic. How do they decide what’s believable? Where do they risk getting on the wrong track? “We want to get a better idea of how they consider the issues and how they make decisions”, she says. Whether or not people participate in the political process depends to a large degree on whether it actually interests them. But that’s not all, says Waldis. “People’s experience of self-efficacy also plays a decisive role”.

You have to be confident enough to speak up during a debate, and you have to be convinced that you actually understand what it’s all about. Young people get their political education these days more and more in digital contexts, but the traditional socialisation contexts also remain crucial: “For adolescents, their family home is their first port of call if they’re unsure about a particular topic and want to get help in how to approach it”.

In the initial phase of their project, Waldis and her team found around 20 young people from across Switzerland who were willing to keep a media diary for several weeks. Then they conducted personal interviews with them. A clear pattern emerged, showing that most of the online information that they had consumed was image-based or comprised short videos. However, the actual content and its level of reflection varied greatly. These young people also preferred very different social media channels, depending on their language region.

Waldis and her team now want to undertake workshops with them to review their data. This is because her team is actually dependent on their expertise when it comes to slang terms and reviews on social media. Waldis and Co. will then carry out a second round of investigations before conducting an overall analysis of their findings in 2027/28, after which they will use them to issue recommendations to schools and youth workers.

Online voting isn’t making progress

The Swiss federal government and the cantons have conducted 300 trials with electronic voting since 2004. After massive criticism on account of security flaws, the Federal Council decided to overhaul the system in 2019 and subject it to further development, primarily to ensure better oversight and greater transparency. In 2023, the federal government approved a resumption of e-voting trials in the cantons of Basel-Stadt, St. Gallen, Thurgau and Graubünden.

But these trials remain restricted in scope. For cantonal votes, a maximum of 30 percent of voters can use e-voting; on a federal level, that figure drops to 10 percent of voters. There is no limit, however, on e-voting among people with disabilities, while Swiss citizens resident abroad are also entitled to use it without restrictions. The federal government is collaborating with the cantons and municipalities to conduct initial trials for electronic signature collection, so-called e-collecting. But the federal referenda in March 2026 revealed a glitch in the canton of Basel-Stadt. So the system remains on shaky ground.

Adding things up until the numbers figure

In a small constituency with four seats in a cantonal parliament, a party would need at least one fifth of all votes cast to secure a single seat. In a larger constituency with 16 seats, however, one seventeenth of the votes would suffice. This means that getting a higher proportion of the vote has no impact in small constituencies. One way to remedy this imbalance is so-called ‘biproportional apportionment’. “It’s designed to try and combine the best of both worlds”, says Karin Manuela Frick, a political scientist.

Its allocation procedure aims to ensure that parties are represented as proportionally as possible across the entire electoral area. But at the same time, it should also safeguard the proportionality of regional representation. There are two steps to the allocation process. The initial step involves determining how many seats each party would get, were there only a single constituency in the canton. In the next step, the seats thus allocated to the parties are then redistributed among the individual constituencies. This requires a series of mathematical optimisations until everything balances out again.

“In cantons like Zurich – whose long historical development has resulted in many constituencies that are small in size – biproportional apportionment is undoubtedly an elegant compromise”, says Frick. In autumn 2025, she was commissioned by the executive of a political party in the Principality of Liechtenstein to examine the pros and cons of the system. A popular initiative calling for its introduction was held in Liechtenstein, only to be rejected.

Frick points to a number of concerns. For example, the procedure can improve the chances of a small party without abandoning the geographical structure of constituencies, but an individual vote is more likely to help a candidate from your chosen party overall than for a candidate from your particular constituency. “What’s more”, she adds, “the model is pretty complex”. This raises the question as to just how far citizens are likely to trust a system that they can’t understand so easily. To be fair, however, she also admits that there probably isn’t a huge interest among the general public in being able to re-calculate the results of an election for verification purposes.

Voting for people, not just parties

Under a preferential system, voters can vote both for a party list and also for specific candidates on it. Switzerland has employed a proportional representation system since 1918 that allows voters to engage in cumulative voting (entering the name of a candidate twice), to strike out the name of a candidate on a pre-printed ballot, and even split their vote (entering candidates’ names on more than one pre-printed ballot list for cantons with more than one seat on the National Council). According to Philipp Harfst, a political scientist at the University of Göttingen, this process of ‘personalising’ national elections has been increasing steadily across Europe since the 1940s. But such preferential voting is implemented in very different ways, depending on the country in question.

Harfst wanted to know how these different methods affect voting behaviour and election results. So he and two colleagues undertook an experiment with a nationally representative sample of Austrian adults. His team replicated the lists from the elections for the Austrian Parliament and then randomly assigned one of three different preferential voting methods equally among their study participants. They found that when a single candidate was able to get up to two points on an electoral list – as is possible, for example, in Luxembourg or in Switzerland’s cumulative voting procedure – then this primarily benefitted the male candidates. Male voters tended to award the maximum number of points to them, leaving female candidates out of the running.

But if voters had the option to award both positive and negative points – a somewhat rare form of preferential voting that was used in the former Soviet Union and is still in use in Latvia today – then this reduces the gender disparity. “Actively deducting a point from someone just because of their gender is probably a step too far for many men”, says Harfst. But a preferential voting system is challenging for political parties because it limits their ability to influence voting outcomes. “A closed list system allows them to ensure that gender balance and proportional regional representation are upheld”.

All the same, Harfst remains convinced that the positive aspects of preferential voting outweigh the negative. It’s also been proven that people perceive a preferential voting system as more democratic. It can even boost voter turnout, as a 2017 study in Spain demonstrated. “That’s remarkable”, says Harfst. “Because voter turnout is something that’s otherwise very difficult to influence”.